Astrocytomas of the adult cerebral hemispheres are among the most deadly of all human tumors. Although much is known about the genetic alterations in malignant gliomas, there is much less information about the initial genetic events leading to the formation of low-grade astrocytomas (LGA). Many LGAs undergo anaplastic progression over months to years, evolving towards GBM. This evolution results in initial phase in tumor formation that is characterized by a small burden of genetic changes on downstream events and tumor behavior has been difficult to study because of the relative rarity of LGA patients, compared to patients with malignant astrocytomas. Analysis of the genetic changes in high-grade astrocytomas have suggested the presence of distinct disease entities that are manifested in different age groups and develop through different genetic pathways. Alterations of p53 and EGFR, for instance are mutually exclusive events in secondary and de novo GBM. Some genetic changes, such as p53, occur at the same frequency across all grades of astrocytoma, suggesting that these genes have a role in initial tumor formation. The investigators hypothesize that the molecular patterns defining specific subsets of gliomas are determined at the time of LGA formation, and that the initial genotype correlates with the rate of anaplastic progression and survival, and will as with subsequent genetic changes. They propose to test this hypothesis in a cohort of LGA patients with the goals of: 1) understanding early events in glioma formation, 2) understanding the molecular basis of anaplastic progression, and 3) gaining the ability to better predict clinical tumor behavior. For this purpose, they propose to obtain a large, multi-institutional, retrospective sample of LGAs, with well-defined clinical follow up, and adequate tissue for histologic and molecular studies. They will characterize specific genetics alterations, and test for correlations between subsets of these alteration and clinical and biological behavior of the tumors. They will examine genetic changes in tumors at the time of anaplastic progression for those cases in which tumor material is available. As a way forward to new hypotheses, they will determine genome-wide changes in subsets of LGA patients with good and poor outcomes using comparative genomic hybridization (CGH).